186 ALASKAN GLACIER STUDIES
the center, as if there were a dragging and a lateral shove there. It explains the lateral spreading of the glaciers and the forward motion of their fronts by the transfer of ice crowded both laterally and forward by the flowage of the lower ice under the pressure of the thrust from behind. It accounts for the uprising of clear ice areas in the Atre-vida and Variegated Glacier bulbs, alluded to on a later page, and for the notable thickening of portions of these bulbs. By this hypothesis it is easy to understand why the breaking of Malaspina Glacier did not extend far westward into the Seward lobe, but was practically confined to that portion of the Malaspina whose ice supply is furnished by the glacier which gave rise to the wave of advance. Equally well does the glacier flood hypothesis explain the abrupt beginning and abrupt ending of the transformation, for the advance would not begin until the thrust was applied, and its effects would terminate soon after the wave had swept on.
Since the glacier flood hypothesis explains all the facts, while other hypotheses fail in significant respects, we believe that it occupies a strong position. There are, however, two objections which will appeal more or less strongly to those who give this hypothesis consideration. One of these is the rapidity with which the breaking extended throughout the glaciers. It is to be noted, however, that this objection would be equally strong against either of the other theories. The rapidity of progress of the advance and breaking, when once begun, is a fact fully established by observation, and whatever theory is proposed in explanation of the change must of necessity assume that the cause was capable of operating with the rapidity observed. So far as we can see it is fully as easy to conceive such a rapid transformation under the glacier flood hypothesis as under any of the other hypotheses. In the absence of more exact knowledge of the behavior of ice in glacier motion, it is not possible to give to this problem the thorough consideration which it doubtless deserves.
If we may assume for glaciers a state of viscous flowage in those parts of the ice under sufficient pressure,—which some will deny,—it seems reasonable to infer that there will be an increase in liquidity under increase in pressure. Our hypothesis is that the application of pressure from the thrust did thus increase the liquidity of the ice, and to a sufficient degree to bring about the phenomena of advance, thickening and breaking observed. This hypothesis is based upon the necessity of such a result from the phenomena observed in the field, which force us to the theory of viscous flowage as postulated under the glacier flood hypothesis. We make no pretention of discussing the physical aspect of the subject, and until there is more knowledge and fuller agreement upon the simplest principles of the behavior of ice in moving glaciers we feel that there would be little profit in such a discussion.
The second objection which some will urge to the glacier flood hypothesis is that already hinted at in the last paragraph, namely the doubt whether glacier ice does under any conditions behave as a viscous body. Upon this subject opinion is divided, and while some believe in viscosity of glacier ice under pressure, others are equally convinced of its impossibility. Much has been written, on this subject both by physicists and by geologists, and since the result of this discussion has been only to lead to a divided opinion, and since the discussions are accessible to all, and known to most students of glacial phenomena, it does not seem necessary to restate or even abstract them here. There are, however, some phenomena of direct observation in the advancing glaciers of Alaska, and some considerations deduced from these observations, which